[mrtg] Inaccuracies on Cisco

[Graham Blake]

Hi there,
This appears to be more of a Cisco problem than an MRTG problem, but it manifested through MRTG graphs. Perhaps someone has a helpful suggestion to explain this, and maybe get around it; My goal here is to have some semi-accurate reporting. We are running some Cisco 2610s w/ 12.0.5 IOS over a frame relay satellite link - pretty high latency, pretty low speed. To accommodate for some packet loss I incresed the queue size on the Cisco, and started seeing some really strange reporting. It *appears* as though the value of "bytes out" for this PVC is increasing by the actual number bytes exiting the port *plus* the incrementing value of "bytes delayed". I have verified this by adding the "bytes in" value for the PVC on the remote end to the "bytes delayed" value on the sending end, and arrive at the value of "bytes out" on the sending end. Very strange. It seems as though the counter is counting each packet that goes into the queue twice. What this means is I am showing a peak rate of 96kB/sec on my 384kbps PVC. Totally weird, unfortunately I don't think I have discovered a wormhole through which to push more data, because the far end is receiving at the expected 384kbps. Has anyone else seen this behaviour before, perhaps have a sound explanation for why the counter on the Cisco seems schizo, and maybe a solution to work around this?
Thanks,
Graham Blake
SSI Micro Network Services
Yellowknife, NT
Canada


--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi